Macro- and Microclimate Conditions May Alter Grapevine Deacclimation: Variation in Thermal Amplitude in Two Contrasting Wine Regions from North and South America

Home , Cuyo (Argentina), Microclimate, Zonda wind

Int J Biometeorol DOI 10.1007/s00484-017-1400-7

ORIGINAL PAPER

Macro- and microclimate conditions may alter grapevine deacclimation: variation in thermal amplitude in two contrasting wine regions from North and South America

Francisco Gonzalez Antivilo1 & Rosalía Cristina Paz2 & Markus Keller3 & Roberto Borgo4 & Jorge Tognetti5,6 & Fidel Roig Juñent1

Received: 21 April 2017 /Revised: 6 June 2017 /Accepted: 16 June 2017 # ISB 2017

Abstract Low temperature is a limiting factor that affects Because local information on meteorological events as prob- vineyard distribution globally. The level of cold hardiness able causes is scarce, this research was designed to test and acquired during the dormant season by Vitis sp. is crucial for study this assumption by comparing macro-, meso-, and mi- winter survival. Most research published on this topic has croclimatic data from Mendoza, Argentina, and eastern been generated beyond 40° N latitude, where daily mean tem- Washington, USA. The goal was to unveil why freezing dam- peratures may attain injurious levels during the dormant sea- age has occurred in both regions, despite the existence of large son resulting in significant damage to vines and buds. climatic differences. Because environmental parameters under Symptoms of cold injury have been identified in Mendoza field conditions may not correspond to data recorded by con- (32–35° S latitude), a Southern Hemisphere wine region char- ventional weather stations, sensors were installed in vineyards acterized by a high thermal amplitude, and warm winds during for comparison. Microclimatic conditions on grapevines were the dormant season. These symptoms have usually been at- also evaluated to assess the most vulnerable portions of field- tributed to drought and/or pathogens, but not to rapid grown grapevines. In order to better understand if it may be deacclimation followed by injurious low temperatures. possible to modify cold hardiness status in a short period with high thermal amplitude conditions, deacclimation was in- Electronic supplementary material The online version of this article duced using a thermal treatment. Hence, despite the fact that (doi:10.1007/s00484-017-1400-7) contains supplementary material, Mendoza is warmer, and temperatures are not as extreme as in which is available to authorized users. Washington, high daily thermal amplitude might be partially involved in plant deacclimation, leading to a differential cold * Francisco Gonzalez Antivilo [email protected] hardiness response.

Keywords Cold hardiness . Deacclimation . Thermal 1 Laboratorio de Dendrocronología e Historia Ambiental, IANIGLA, amplitude . Grapevine . Mendoza . Washington state CCT-CONICET-Mendoza, Av. Ruiz Leal s/n, Parque Gral. San Martín,, PO Box 5500, CC 330 Mendoza, Argentina 2 CIGEOBIO (FCEFyN, UNSJ/CONICET), Av. Ignacio de la Roza 590 (Oeste), J5402DCS, Rivadavia, San Juan, Argentina Introduction 3 Department of Horticulture, Irrigated Agriculture Research and Extension Center, Washington State University, Prosser, WA 99350, Climate affects the distribution of vineyards, resulting in lo- USA calization of the main areas of viticulture between 30 and 50° – 4 Cátedra de Fisiología Vegetal, Facultad de Ciencias Agrarias, Nand30 40° S. These latitudes define areas corresponding Universidad Nacional de Cuyo, Almirante Brown 500, Luján de approximately to isotherms between 10 and 20 °C. As a pe- Cuyo, Mendoza, Argentina rennial temperate plant, the physiology of this climbing plant 5 Laboratorio de Fisiología Vegetal—Facultad de Ciencias Agrarias, is adapted to a marked seasonality modulated by temperature. Universidad Nacional de Mar del Plata, Mar del Plata, Argentina Thus, warm conditions during the growing season (GS, spring 6 Comisión de Investigaciones Científicas de la Provincia de Buenos to autumn) are crucial for plant development and fruit produc- Aires, Buenos Aires, Argentina tion, while during autumn-winter, the cessation of growth Int J Biometeorol leads to a dormant season (DS). During the DS, air tempera- located in maritime regions (e.g., New Zealand, Chile, ture usually falls below the freezing point, causing damages to Australia, and South Africa) with low incidence of freezing vulnerable tissues. This situation is one of the main causes of damage (Mullins et al. 1992). Therefore, there are scarce re- yield losses and/or plant death in grapevines (Fennell 2004). gional data concerning CH in this hemisphere, as regards Two factors, plant cold hardiness (CH) and temperature, vineyards or any other temperate fruit crops (Supplementary interact to determine tissue damage by freezing. Most research Fig. 1). Mendoza Province (MZA), the main grape producing has focused on CH, since this phenomenon constitutes a dy- region of Argentina, is an exception to this environmental namic and complex trait acquired in response to a shortening condition, as evidenced by a high level of continentality and photoperiod and declining temperature in autumn (Howell altitude (Table 1). and Shaulis 1980; Fuchigami et al. 1982; Wisniewski et al. Typical symptoms of cold injury include lack of and/or 1996). Three different stages are identified for CH: (i) accli- uneven bud break, death of canopy, under soil plant re- mation, when the plants gain CH; (ii) deacclimation, when the sprouting, and trunk cracking (Brusky Odneal 1983; plants lose CH; and (iii) re-acclimation, when the plants regain Goffinet 2004 ). In Mendoza, these symptoms have also been CH after temporary deacclimation (Levitt 1980). Rates of ac- observed but have been attributed to drought and pathogen climation and deacclimation vary dynamically during the DS, attack because freezing damage is not considered to be a crit- and are reversible (Dambrorská 1978; Wolf and Cook 1992; ical factor in vineyards. This occurs firstly, because there is Gu et al. 2002). CH is a reversible process; therefore, a cold little available information about the range of CH in grapevine hardy plant can deacclimate and then re-acclimate depending cultivars from local vineyards, which are usually extrapolated on temperatures. This process differs with species, cultivar, from studies conducted in the Northern Hemisphere. phenology, organ, weather (Xin and Browse 2000;Gusta Secondly, most of the microclimatic field conditions are esti- and Wisniewski 2013; Pagter and Arora 2013), and crop man- mated by extrapolation from data generated in CWS and not agement (Wample and Wolf 1996). from temperature sensors installed at a plant level. Finally, Temperature, the second factor, is a complex parameter and variability among years, in winter temperature, makes it diffi- is highly variable in time and space, affecting biological sys- cult to associate the dormant with the symptoms of cold dam- tems directly (Eagles 1989; Sage and Kubien 2007). This fac- age by producers and agronomist. tor can be considered at different spatial scales: macroclimate (a regional scale of tens to hundreds of kilometers), Table 1 Comparison of climatic and geographical condition between mesoclimate (a vineyard scale of tens to hundreds of meters), MZA and WA and microclimate (the specific environment around any plant) (Robinson 2006). Traditionally, the following limited number Item MZA WA of parameters has been measured to estimate thermal condi- Latitude (°) 32–35 S 46–49 N tion at the three mentioned scales: (i) maximum temperature asl (m) 766 117 (Tmax), (ii) minimum temperature (Tmin), (iii) daily mean Köppen-Geiger BW.k—arid, desert, BS.k—arid, steppe, temperature (Tmean), (iv) daily thermal amplitude (TA), and classification cold arid cold arid (v) relative humidity (RH). These parameters are traditionally Annual precipitation 213 204 obtained from conventional weather stations (CWS), which (mm) according to international standards consist in a white woody Pluvial regime Summer Winter shelter located at 1.4 m above the soil and sensors confined Annual temperature 16.4 12 mostly under the shield to protect them from the influence of average (°C) Sea of influence Pacific Pacific precipitation and direct radiation. Moreover, these stations are Range of influence Los Andes Cascade often located at a considerable distance from a vineyard. This Wind condition Lee Lee combination of elements often leads to under- or overestima- Mean altitude of range 4000 1300 tion of the real thermal environment to which a plant is sub- (m)* jected to which may result in misinterpretation and/or minimi- Width (km)* 150 70 zation of the influence of temperature on plant damage. Distances foothills to 20 130 In the Northern Hemisphere, most vineyards beyond 40° N vineyards (km) are affected by recurrent climatic contingencies due to freez- Continuity Continuous Discontinuous ing damage. This implies a direct impact on the growth man- Continentality index Continental climate Slightly continental (Ivanov) (165.3) climate (122.7) agement (Mills et al. 2006; Cragin et al. 2017;Hammanetal. Continentality index 26.9 10.6 1996; Fennell 2004; Keller and Mills 2007), and a permanent (Conrad; %) real-time monitoring and measurement of CH (http://wine. wsu.edu/research-extension/weather/cold-hardiness). On the Sources: Climatic-data.org, 2017; Google Earth, 2017 other hand, in the Southern Hemisphere, most vineyards are *measurement at the study region Int J Biometeorol

MZA is located at the center-western side of the country corresponding to April–May in SH and October–November near the foothills of the Andes Range, whereas the wine grow- in NH, middle DS corresponding to June–July in SH and ing region of eastern Washington (WA), USA, is an area lo- December–January in NH (around the winter solstice), and cated between the Cascade Ranges to the west and the Rocky late DS corresponding to August–September in SH and Mountain Range to the east (Table 1). Both regions show February–March in NH. orographic similarities and are subject to periodic warm winds from the Pacific Ocean locally named Zonda in Argentina and Comparison between temperature measurements Chinook in the USA. These winds are strong and warm and in the field and CWS are associated with adiabatic compression upon descending the eastern slopes of mountain ranges (Norte and Simonelli The characterization of field microclimatic conditions was 2016). Both, Zonda and Chinook, mainly occur during the done in vineyards belonging to the Plant Physiology DS, provoking unseasonal warming periods. Moreover, after Department of the Faculty of Agricultural Sciences the windy days are often followed by hard freezing events. (UNCuyo) in Luján de Cuyo, MZA (33° 0′ 29 60″ S latitude; However, it is not clear how these atemporal warming events 68° 52′ 21″ W longitude) and to the IAREC in Prosser, WA influence CH. (46.2° 15′ 25″ Nlatitude,11944′ 25″ ° W longitude) during This research aimed to compare climate conditions be- the DS of 2013 and 2012, respectively. tween MZA and WA during the DS in order to understand To determine temperature differences between field and differences and similarities and unveil why there is freezing CWS from both locations, the same equipment was used to damage in both regions despite their differences. This com- record field data (Thermochron DS1922L-F5 iButton temper- parison was performed at three scales: (i) macroclimate, using ature loggers, Maxim Integrated, San Jose, CA, with a mea- the 10-year climate data recorded from CWS (Tmax, Tmin, surement range of −40 to +125 °C, and accuracy ±0.5 °C). At HR, and TA); (ii) mesoclimate, evaluated by comparing daily each vineyard site, three sensors were installed adjacent to field temperature measurements with data recorded by CWS grapevine trunks at 40 cm above the soil and protected from in order to evaluate its reliability and accuracy to predict the solar radiation by reflector panels. Temperatures were record- real field conditions; and (iii) microclimate, evaluated at plant ed hourly during February–March 2012 in WA, and during level, to determine the most vulnerable portions of the plant to July–August 2013 in MZA and used to calculate daily Tmin damage. The second objective was to evaluate how TA during and Tmax. Data were recorded for at least 15 days in the the DS may determine loss of CH in cold-acclimated plants. middle DS, and repeated three times in each location. To com- For this purpose, short-term experiments with acclimated pare temperatures at field conditions with those registered plant exposed to different TA treatment for 24 h were per- from CWS located nearby the vineyards (500 m approximate- formed in order to measure the rate of deacclimation following ly), daily maximum and minimum temperatures for the same exposure of grapevine to different TA. time period were also obtained. In the case of vineyards from MZA, data were obtained from the weather station of the Faculty of Agricultural Sciences (33° 00′ 18″ S latitude; 68° Materials and methods 52′ 13″ W longitude) whereas for those from WA, the weather station was the IAREC Headquarter described previously. Macroclimate conditions in MZA and WA Plant microclimatic conditions The meteorological conditions during the vineyard DS (autumn-winter period) at the regions of MZA and WA were In order to characterize microclimatic conditions close to evaluated for 10 years (2002–2012) using data collected in grapevines during the DS, three independent assays were con- two official weather stations: a DACC (Direction of Climate ducted, as indicated in Fig. 1: (a) air thermal differences be-

Contingencies) station located at INTA La Consulta, San tween the top (Ttop) and the base (Tbase) of the plant were Carlos, Mendoza Province, Argentina (33° 43′ 10″ Slat;69 measured with sensors located at 1 and 0.1 m above the soil, 69° 0.6′ 16″ W long) and a Washington State University, respectively; (b) above the soil versus below the soil thermal AgWeatherNet (http://weather.wsu.edu), station located at differences with sensors installed near the plant at 0.1 m above the Irrigated Agriculture Research and Extension Center the soil (Tbase) and 0.1 m below soil (Tsoil); and (c) internal (IAREC) Headquarter, Washington, USA (46° 15′ 25″ N; trunk temperature (Tin) versus air temperature near the plant lat, 119 44′ 25″ W long). To standardize data from Northern (Tout), with sensors installed 0.1 m above the soil, one inside (NH) and Southern Hemispheres (SH), the beginning of au- the trunk and the other placed at 0.05 m from the plant. tumn was considered as day 1 of the DS. Daily parameters Because iButtons could not be placed in the trunk, thermo- evaluated in this study were Tmax, Tmin, TA, and RH. DS couples were set up in the internal portion of the trunk by was broken down into three different stages: early DS drilling a hole of 1 mm in diameter and 10 mm deep. Then, Int J Biometeorol

buds 4 through 8 were used and divided into buds and internode sections, which were mixed and randomized for freezing treatments (n = 20 for buds section and m = 8 for internodes). Each cane was split into two parts: the bud and the internode sections. The bud section consisted in the bud itself and nearly 2 mm of subtending cane tissue above and below the bud (Andrews et al. 1984; Wolf and Pool 1987). The internode section of the cane consisted in 2-cm-long cane tissue without the bud section. Plant material was obtained after sunrise (9:00 a.m.). Canes were placed in plastic bags with a wet paper towel inside to prevent dehydration. Also, during both trans- port and preparation of samples in the laboratory, temperatures were maintained slightly above 0 °C to avoid heating. Three treatments simulating different temperature changes prior to freezing conditions were applied to determine the deacclimation temperature of buds and internode sections. In all treatments, the temperature at the sampling time was around 0 °C. The treatments were (a) control: sections without any thermal treatment and evaluated after collection; (b) low TA: sections maintained at 7 °C in a fridge for 24 h; and (c) Fig. 1 Grapevine microclimate characterization during the dormant high TA, maintained at 30 °C in a heated cabinet for 24 h. season in Mendoza, Argentina (MZA) and Washington, USA (WA). Following thesis treatment, CH of grapevine buds and cane Three independent assays were conducted to determine (a) air thermal tissues from each treatment was estimated by the differential differences between the top (Ttop) and the base (Tbase) of the plant, with sensors (filled circle) located at 1 and 0.1 m, respectively; (b) soil versus thermal analysis (DTA) method, based on the low-temperature air thermal differences with sensors (filled square) installed near the plant exotherms (LTE). This method was showed by Mills et al. − at 0.1 m above the soil (Tbase)and 0.1 m below soil (Tsoil); (c) internal (2006) and correlated very closely with visual damage of cane trunk temperature (Tin) versus air temperature near the plant (Tout)with sensors installed 0.1 m above the soil, one inside the trunk and the other at phloem and xylem, based on the browning tissue. Briefly, the a distance of 0.05 m from the plant DTA consists of a chamber of thermoelectric modules (TEM) integrated with a commercially available programmable freezer and data acquisition system (DAS). TEM consist in 40 a T-type thermocouple (copper-constantan; measurement channels of Peltier plate model CP1.4-127-045L (Melcor range −40 to +125 °C, accuracy ±0.1 °C) was inserted at the Corporation, Trenton, NJ) that senses exotherms produced level of the trunk phloem (5 mm deep) and then covered with when water or tissues freeze in each channel and converts this tape. Thermocouple data were collected by data loggers of thermal signals at an output voltage (mV). The programmable different trademarks but with the same functions: Cava freezer model T2C (Thermal Product Solutions, Williamsport, Devices, four channels in Argentina and Campbell Scientific PA) was equipped with a temperature controller model 44212; CR10 in USA. All temperature measurements were done in (YSI, Dayton, OH) that senses freezer temperatures. All mea- three different adjacent plants of Vitis vinifera cv. Malbec in surements from individual TEM channels and temperature MZA and Vitis labrusca cv. Concord in WA. Plants of each were acquired and recorded at 15-s intervals in DAS, that cultivar had trunks of similar diameter (between 4 and 5 cm) consist in a multi-parameter data logger Keithley model and were of similar age (8 to 10 years old). Daily measure- 2700-DAQ-440 (Keithley Instrument, Cleveland, OH). The ments were divided into two parts, day and night, and daytime low-temperature exotherms corresponding to 10% phloem measurements were taken from 8:30 a.m. to 7:00 p.m. in MZA injury and 10% xylem injury, respectively, were determined and from 7:30 a.m. to 5 p.m. in WA. as shown in Supplementary Fig. 2. The DTA data for grapevine bud and internode sections had previously been shown to Effect of high TA on deacclimation of buds and cane correlate closely with visual damage as determined by tissue tissues browning (Mills et al. 2006). Samples were placed directly on TEM chambers. From For this experiment, 12-year-old field-grown grapevines Vitis each treatment, 20 buds were placed with the cut surface fac- vinifera cv. Chardonnay and cv. Merlot were sampled in the ing up; whereas in the case of cane tissues, eight of such cane IAREC vineyards. Twenty plants per cultivar were selected, pieces were placed directly on each TEM. The rate of descend and we cut off one full cane from each above the fourth node for freezing simulation was the standard for DTA method to preserve the more basal nodes for spur pruning. Only the being 4 °C/h (Mills et al. 2006). It was demonstrated that this Int J Biometeorol descend rate is statistically similar over cold injuries if com- early, middle, and late DS, respectively). Major differences pared with the real rates observed in field (Haynes et al. 1992). were observed during the middle DS (Fig. 2a).

Lethal temperatures for buds were reported as bud LT10, MZA was relatively constant throughout all the DS, oscil- (i.e., lethal temperatures at which 10% of the buds are killed; lating around 60% (characterized by a dry winter season). On Andrews et al. 1984). For internode sections, lethal tempera- the other hand, RH in WA started to increase during the early tures were reported as phloem LT10 and xylem LT10. DS peaked at almost 90% in the middle DS (characterized by high frequency of rains and snowfall) to decrease gradually Statistical analysis during late DS until reaching similar values than those registered in MZA (Fig. 2b). The statistical analysis was performed with R (R Core Team The seasonal progression of TA was closely related to that 2013) and Infostat software v. 2016 (Di Rienzo et al. 2016). of RH (Fig. 2c). Consequently, in MZA, the TAwas relatively For the analysis of macroclimatic conditions of the two con- constant throughout the DS (around 16 °C) but with a signif- trasting locations (MZA and WA), a generalized linear mixed icant higher interannual variation (MZAvar(n − 1) =27.6°Cand model (GLMM) method was used. The final model included WAvar(n − 1) =11.3°C;p <0.0001).UnlikeinMZA,theTAin the standardized day of year and location as fixed effect, and WA showed a gradual decline during early DS, reaching min- year as a random effect. Data were partitioned according to the imum values of nearly 6 °C during the middle DS, to continue DS subperiod (early DS, middle DS, and late DS). Significant with a gradual increase during the late DS. differences between mean values of Tmax, Tmin, TA, and RH were determined by the DGC test with α <0.05(DiRienzo Comparison between temperature measurements et al. 2002). in vineyard and CWS To compare data obtained from field sensors against data from CWS, a paired t test with α < 0.05 was used. The daily Close correlations were found between data obtained from differences between the weather station and the field sensors field sensors and CWS in both MZA and WA (MZA: 2 of Tmax (Tmax diff = Tmaxstation − Tmaxfield)andTmin Tmaxfield vs. Tmaxstation r =0.91andTminfield vs. 2 2 (Tmin diff = Tminstation − Tminfield) were estimated. The same TminStation r =0.92;WA:Tmaxfield vs. Tmaxstation r =0.90 2 analysis was applied for TA. and Tminfield vs. Tminstation r = 0.92). However, at both sites, Daily Tmax, Tmin, and TA measured in the microclimatic the TA registered close to the plants (field conditions) was analysis were analyzed with the same methodology employed significantly higher than that registered by the CWS (Fig. 3). to analyze macroclimatic conditions. The time (in minutes) In WA, both field TA and CWS TA were lower than in MZA. was considered as a fixed effect whereas temperature mea- The Tmax and Tmin differences registered between the sured by sensors was considered as a random effect, and the field and the CWS revealed that in none of the cases that the data were partitioned in day-night values. Moreover, in the difference the TA was equal to zero. However, the CWS analysis of Tin versus Tout gradient, the average day was esti- tended to underestimate Tmax and overestimate Tmin (Fig. mated using a generalized linear model (GLM). 3). Field plants were exposed to higher Tmax values than For the analysis of differences in CH between tissues, treat- those recorded by CWS: mean Tmax were 3.04 and 2.52 °C ments, and cultivars, repeated measures one-way ANOVA higher in the field than in MZA and WA weather stations, were used. When effects were significant, multiple compari- respectively. Also, mean values of Tmin were 1.70 and sons were performed using Duncan’s post hoc test. 0.71 °C lower in the field than CWS values from MZA and WA, respectively. It is important to note that rather large differences (e.g., 10 °C) were sometimes found between field Results and CWS-recorded temperatures (circled symbols in Fig. 3).

Comparison of macroclimatic conditions between MZA Plant microclimatic conditions and WA The analysis of plant microclimatic conditions close to the The average of daily temperature during the DS in the 10-year base of the trunk (10 cm above the soil level) and near the period analyzed was higher in MZA (9.8 °C) than in WA top (100 cm above the soil level) of dormant grapevine cano- (5.2 °C). Moreover, average Tmax was always higher MZA pies revealed that during the day, there was no vertical tem- than in WA (TmaxMZA-TmaxWA are 6.2, 11.3, and 5.8 °C perature gradient, while at night, the base of the plant is expe- during the early, middle and late DS, respectively). By con- rienced to lower temperatures, a situation that leads to a higher trast, Tmin was relatively similar at the two locations and only TA at the base than at the top of the plant (Table 2(A)). This showed significant differences during the middle DS pattern was similar in both MZA and WA vineyards.

(TminMZA-TminWA are −2.1, −1.2, and −0.1 °C during the However, differences were larger in MZA, with a night Int J Biometeorol temperature gradient of −2.03 °C between base and top, and a Fig. 2 Ten years of macroclimatic records (2002–2012). Mendoza, TA difference of 3.93 °C, compared to a night temperature Argentina (MZA; 33° 43′ 10″ S latitude; 69° 0.6′ 16″ W longitude) and ′ ″ ′ ″ gradient of −0.22 °C and TA differences of about 1.25 °C in Washington, USA (WA;46° 15 25 N latitude, 119 44 25 W longitude). (a) Comparison of daily maximum(Tmax)andminimum(Tmin) the WA vineyard. temperatures, (b) relative humidity (RH), and (c) thermal amplitude Soil temperatures were also significantly different from (TA) during the early, middle, and late dormant season of grapevines with the temperatures registered at the base of the plant (October–MarchinUSAandApril–September in Argentina). (Table 2(B)). In this case, similar results were obtained in Significant differences between mean values were determined by the DGC test with α <0.05(n = 10 years) MZA and WA, with the soil being colder than the air near the base during the day (Tbase-Tsoil = −4.87 and −4.49 °C in MZA and WA, respectively) and warmer during the night have been generated in order to predict cold damage during

(Tbase-Tsoil = 2.71 and 3.13 °C in MZA and WA, respectively). the DS (Ferguson et al. 2011, 2014). These models were ap- A similar effect was also observed in TA, where the soil ex- plied in regions where the temperature remains relatively low hibited less significant diurnal temperature variation than the and stable during DS (Proebsting et al. 1980;Jiangand air near the trunk base (TAbase-TAsoil = −11.82 and −10.84 °C Howell 2002). However, the same models did not perform in MZA and WA, respectively). well under the climatic conditions of MZA (Gonzalez Significant differences were also observed between the Antivilo; unpublished results), suggesting that the model does temperatures inside (Tin) and outside (Tout) the vine trunk with not capture important aspects of climatic variation among a similar response at both sites (Fig. 4). Whereas Tin was growing regions and the importance to generate local infor- slightly higher than Tout during the night, during the daytime, mation and fine CH monitoring in order to understand the the trunk was warmer than the air. This nighttime difference causes of freezing damage in this region. was 1.3 °C in MZA and 2.7 °C in WA. On the other hand, was Both MZA and WA are viticulture regions of global impor- observed that at sunrise, the air temperature of the surrounding tance and share some climatic similarities, grape cultivars, as air increased faster than the trunk, but at midday, the temper- well as irrigation, diseases, and other vineyard management. ature of the trunk was greater than of the air. This pattern is Geographically, both sites have in common the proximity to maintained until around 8:00 p.m., the moment from which mountains, episodes of Föhn wind, similar bioclimatic index-

Tin and Tout became similar. es according to Köppen-Geiger (cold, arid, and desert), and annual precipitation around 200 mm (Table 1). Also, both Effect of TA on deacclimation rates of buds and cane regions show signs of frost damage. However, a comparison tissues of 10-year meteorological records from both regions revealed contrasting differences with potential influence in the physi- In order to determine whether tissue deacclimation may be ology of Vitis sp. plants. induced by differences in TA, field acclimated cane internodes The use of Tmean to compare climatic characteristics and buds were subjected to Low-TA (7 °C), High-TA (30 °C), among regions and its association with biological parameters and control (0 °C) during 24 h of artificial thermal treatments (i.e., photosynthesis, CH) are widely accepted. According to in two cultivars. this bioclimatic index, the DS of MZA is ~5 °C warmer than

The LT10 of Chardonnay buds and internode tissues did not that of WA in. However, deepening the remainder parameters differ between the control and Low-TA. Conversely, there was that define thermal characteristics, it is possible to generate a significant effect of the High-TA treatment. The tissue that another interpretation during the DS from both sites: (i) de- was most vulnerable to deacclimation was the internode xy- spite latitude differences, there are slight differences in aver- lem, followed by the phloem, and finally the bud (Table 3). age Tmin (<1.5 °C); (ii) mean Tmax is considerably higher in In Merlot, neither buds nor xylem buds and xylem showed MZA than in WA, reaching maximum differences during the significant differences between any treatments. However, middle DS with mean values of 12 °C and often higher daily High-TAledtolossofphloemCHincv.Merlot(Fig.5, differences; (iii) the average TA in MZA is ~16 °C with high Table 3). interannual variation throughout the whole DS period, whereas in WA, the TA reaches a minimum during the middle DS (~6 °C); (iv) moreover, in WA,but not in MZA, there are some Discussion days when Tmax did not exceed 0 °C (Fig. 2). Considering the vulnerability of V. vinifera to deacclimation induced by Most information concerning CH in grapevines in MZA is unseasonal high temperatures (Ferguson et al. 2011, 2014), currently extrapolated from information generated from dif- these climatic differences could lead to occasional cold dam- ferent regions of the NH, because there is no local informa- age even in MZA vineyards. tion. In WA, however, cold injury is a common phenomenon The presence of the mountain barriers produces a consid- with severe economic consequences. In fact, different models erable influence on the regional climate, deeply influencing Int J Biometeorol Int J Biometeorol

Fig. 3 Comparison of ab 30 30 microclimate variation between TAStaon: 15.88°C TAStaon: 8.82°C field conditions and official 26 TAField: 20.83°C 26 TAField: 12.05°C p < 0.0001 p < 0.0001

weather station in Mendoza, 23 C) 23 ° Argentina (MZA; left)and 19 19 Washington, USA (WA, right). 15 ature ( 15 (a)and(b) show the daily thermal r 11 11 amplitude variation measured in Tempe Temperature (°C) Temperature 8 8 the field (empty circles) and by the weather stations (filled 4 4 0 0 squares). (c) and (d)show 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 differences in Tmax (filled Day of year Day of year triangles)andTmin(filled inverted triangles)betweenthe cd6 Mean Tmax diff: -3.04 (p<0.0001) 6 Mean Tmax diff: -2.52 (p<0.0001) weather station and the field Mean Tmin diff: 1.70 (p<0.0001) Mean Tmin diff: 0.71 (p<0.0005) ) 5 5 ld − ld) (Tdiff = TStation TField). Mean e ie 3 i

F 3 F ° values of thermal amplitude, ° -T 2 -T

n 2 Tmax diff, and Tmin diff are also n io io t t a shown. Circled symbols indicate 0 a 0 St

extreme values. Measurements T° -2 -2 ( (T°St

f were conducted during the f di

diff -3 -3

dormantseasonof2013and2012 ° T in MZA and WA, respectively -5 T°C -5 -6 -6 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Day of year Day of year seasonal TA, precipitation, and continentality (Thompson Both MZA and WA regions are affected by two semi- et al. 1977). However, the magnitude of the effect caused by permanent high-pressure centers located above the South these barriers depends on the orographic characteristics and North Pacific Oceans, respectively. In the coastal regions in each case. In this sense, the Andes consists in a for- (windward side), these pressure centers are responsible for wet midable obstacle to the Pacific Ocean winds, consisting winters and dry summers (Mass and Dotson 2010; in more than threefold of mean altitude and the double of Ancapichun and Garcés-Vargas 2015). During DS, those an- width than the Cascade Range (Table 1). Moreover, ticyclones push winds toward the mountain ranges. As the whereas the Andes is an uninterrupted mountain system, wind climbs the slopes of the windward side of the range, theCascadeRangeisbrokenupbytheColumbiaRiver the moist air loses water as rain or snow. However, the altitude gorge, showing a lower continentality index than MZA of mountain ranges is able to modify the rainfall distribution (Gedalof et al. 2005; WRCC 2013). on both sides of the mountains (Bianchi and Yáñez 1992). In

Table 2 Plant microclimate characterization in MZA and WA during the DS

MZA WA

Day Night TA Day Night TA

(A) Vertical gradient in air

Ttop 12.6 ± 0.3 3.5 ± 0.2 17.1 ± 0.4 6.0 ± 0.1 1.6 ± 0.1 10.7 ± 0.5

Tbase 13.2 ± 0.3 1.5 ± 0.2 21.0 ± 0.4 6.6 ± 0.1 1.4 ± 0.1 11.9 ± 0.5 r2 0.61 0.81 0.31 0.68 0.72 0.08 p ns <0.0001 <0.0001 ns 0.035 ns (B) Vertical gradient in soil

Tbase 7.2 ± 0.24 4.8 ± 0.09 7.4 ± 0.63 7.4 ± 0.14 5.8 ± 0.27 6.5 ± 0.94

Tsoil 12.1 ± 0.24 2.1 ± 0.9 19.2 ± 0.63 11.9 ± 0.14 2.7 ± 0.27 17.3 ± 0.94 r2 0.65 0.84 0.79 0.73 0.68 0.65 p <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

(A) Day and night air thermal differences and TA between top (Ttop) and base (Tbase) of trunk. (B) Day and night air thermal differences and TA between above ground (Tbase)andbelowsoil(Tsoil). ns: not significative Int J Biometeorol

Fig. 4 Hourly microclimate characterization inside and outside (a)) and Washington, USA (WA; (b)). Symbols are means and bars are grapevine trunks during the dormant season. Diurnal temperature standard deviation (n = three plants) measured during 15 days in August change of trunk (Tin; filled diamond) and air near the plant (Tbase;) in an 2013 in MZA and during 6 days in February 2012 in WA average day during the dormant season in Mendoza, Argentina (MZA; the case of the Andes, its high altitude (~4000 m asl in aver- 1940).Therefore,theAndesareamuchmoreimportantcli- age) favors rain and snow accumulation at the windward side, matic barrier for MZA (Bianchi and Yáñez 1992), than are the drying the air and preventing precipitation in the low plains of Cascades for WA, where their effects are less marked and show MZA (Bianchi and Yáñez 1992). However, the relative lower more influences by the Pacific Ocean (Mass and Dotson 2010). altitude of the Cascades (~1000 m asl on average) and their Temperature and RH are both variables closely related; interruption by the Columbia river allow that clouds surpass consequently and depending on the degree of humidity of the hills in their drift toward the west, favoring winter rains on the air is the amount of energy that is required to heat or cool both sides of the mountains (Mass and Dotson 2010). As a it, needing a saturated air of moisture more energy to raise its consequence of these characteristics, clear sky is frequent temperature (specific heat saturated air 0.43 kcal/kg/°C) than a (~100%) in MZA during the middle DS, whereas in WA, the frequency of sunny days is significantly lower (~25%). Another consequence of the passage of winds across the mountains is the BFöhn effect.^ This phenomenon is caused by the adiabatic compression that winds suffer when they descend the lee slopes, causing a temperature increment and an unseasonal Blittle summer in winter^ followed by a strong freezing event. As more pronounced is the altitude, as more strong, warm, and dry is the wind (Norte and Simonelli 2016). In the case of the Zonda winds, common in MZA, their occur- rence is distributed from May to November, with nearly 50% of their incidence within the July–August months (late DS). Moreover, the probability of this wind would be severe (>10 °C) and duration (up to 3 days) is only high in late DS (Caretta et al. 2004). By contrast, the Chinook winds in WA rarely occur during the winter, and produce a relative small thermal effect due the low altitude of the Cascades (Burrows

Fig. 5 Lethal temperatures (LT ) in grapevine bud and cane tissues Table 3 Average difference between LT10 and LT10 10 30 °C control following different thermal treatments prior to a freezing simulation in Tissue Chardonnay Merlot cv. Chardonnay and cv. Merlot. Buds and internode sections were submitted to two temperatures of low thermal amplitude (7 °C TA) Bud −1.54 ns (filled circle) and high thermal amplitude (30 °C TA) (filled triangle) for 24 h followed by controlled freezing. The controls (filled square) − − Phloem 3.55 2.62 were left untreated before controlled freezing. Symbols are means and Xylem −4.43 ns bars are standard deviations (internode: n =8;buds:n =20).Symbols with different letters within a column are significantly different (Duncan’s ns: not significative test, p < 0.0001) Int J Biometeorol totally dry air (specific heat dry air 0.24 kcal/kg/°C). In MZA, macroclimatic conditions of a particular region, our data sug- HR is quite constant during the DS (~60%) and with low gest that to reach more accurate biological inferences, it is probability of precipitation as rain or snow, resulting in with recommended to set up in situ sensors in order to understand a very high TA (Fig. 2b, c). Moreover, Zonda winds have a real temperature conditions to which plants are exposed. high desiccant power that reduces HR at levels lower than Moreover, this control scheme has direct implications in ac- 30% (Norte and Simonelli 2016) being probably the main tive control, monitoring during freezing events, and evalua- cause of the high interannual variation in daily TA observed tion of damage after it occurs. in this research (Fig. 2c). Conversely, in WA, the increase in At individual, the plant scale, there is spatial variability in HR shows a progressive increment reaching a highest value at freezing resistance and risk, from the root system to the top of the middle DS (~100%), a fact that explains the low TA ob- the canopy (Charrier and Ameglio 2011, 2015). It is widely served there. In fact, on some days, the maximum and mini- accepted that there is a vertical air temperature gradient, and mum temperatures are almost equal (TA ≈0). In this sense, our the lowest temperature is close to soil (Leuning and Cremer results suggest that the contrasting peak between climat- 1988;JordanandSmith1995; Blennow 1998; Battany et al. ic situations between MZA and WA coincides with the 2012). However there is scarce information about the magni- late DS, a period when grapevines are most susceptible tude of this gradient and its implications on acclimation and to deacclimation. deacclimation processes during the DS. Our results suggest Air temperature is recorded from different types of meteo- that in MZA, the base is subject to high TA (Table 2(A)). rological weather station and sensors. One of the most tradi- This phenomenon could potentially lead to deacclimation, tional and more employed worldwide are the Stevenson leaving the base of the trunk to more exposed conditions to weather stations, created in the XIX century and composed cold injury during an unusually cold night. According to our by different analogical sensors protected by a white woody observations, damage to the trunk base is the most common shelter. Nowadays, those stations are being replaced by mod- symptom of cold injury in grapevine checked in MZA; there- ern digital automatic weather stations with white plastic shel- fore, it could be related with a differential deacclimation be- ter. However, temperature is a highly variable and sensible tween different parts of trunk of grapevine. Conversely, in parameter and it depends on the sensors employed, construc- WA, the near absence of this temperature gradient might be tive characteristics of shelter, closeness to buildings, etc. attributed to the higher RH during winter. (Tarara and Hoheisel 2007;Falletal.2011; Menne et al. Trunk temperatures at the base level also had higher Tmax 2010). In fact, small differences such as the kind of employed and consequently higher TA than the surrounding air. The paint affect the real temperature values of Tmax in 1.66 °C trunk’scapacitytogainenergyisrelatedtoitsmass,which (0.3 °F) and Tmin in 2.78 °C (0.8 °F) (Fall et al. 2011). generates thermal inertia so that trunks start to warm up or The literature refers that thermal data recorded by different cool down later than does the air (Fig. 4). A similar result sensor technologies and shelter materials depend on the geo- has been observed by Paroschy et al. (1980). Although trunk graphical scope, type of sensors, and researcher preferences. temperatures are generally higher in MZA than in WA, the Meteorological research is commonly based on data generated presence of snow in WA may produce an albedo effect, induc- by CWS (Giantomasi et al. 2009; Burgos 2010). However, in ing a major thermal gain during daytime (Fig. 4). biological studies relating observations of living organisms Although the root is the most susceptible organ to freezing with weather parameters, it is necessary to employ data ob- damage (Okamoto et al. 2000), the soil has thermal inertia that tained by nearby CWS and/or by data supplemented by sen- acts as a buffer (Table 2(B)). This capacity protects the root sors installed in situ (Hubackova 1996; Echarte et al. 2010). from excessive temperature decreases. Moreover, it allows The greater is the precision required for an inference, the shoot re-growth from latent buds on the below ground portion more important is sensor density and proximity to the living of the trunk, a characteristic symptom in freeze injured plants organism under study. In this case, the comparison of temper- (Zabadal et al. 2007). ature records from sensors installed in vineyard and those Few research estimates deacclimation rate by testing differ- from the nearest CWS (<500 m) in MZA and WA revealed ent combinations of temperature and time periods. These that despite field and CWS data being correlated, in both sources determine that it depends on species, varieties, and cases, we observed a significant bias for the CWS to overes- temperature to which the plant is subjected (Pagter and timate Tmax and underestimate Tmin. Thus, the TA value is Williams 2011; Eagles 1989). Thus, in Vitis sp., it was dem- typically higher in the field than indicated by the CWS. onstrated that thermal treatments of 14 °C during 3 weeks (cv. Moreover, in particular cases, the values recorded by CWS Thompson Seedless; Rubio et al. 2016), and 25/20 °C day/ differed markedly from the field measurement (Fig. 3). night during 4 days (cv. Chardonnay; Cragin et al. 2017), Similar results were also obtained by monitoring other plant induced similar daily deacclimation rates of 2 °C. Our results species (Renaud and Rebetez 2009). Thus, although that tem- indicate that high TA may lead to similar deacclimation rates perature records from CWS are useful for describing in 1 day, but at higher TA values. Thus, a TA of 30 °C was Int J Biometeorol strong enough to change the CH status in all tissues and both than that registered by CWS. Moreover, we demonstrate than cultivars evaluated, whereas a TA of 7 °C did not alter CH the thermal environment around individual plants is also var- (Fig. 5). Although Chardonnay had higher initial CH, it was iable. The base of grapevine trunks proved to be the most more susceptible to deacclimation than the seemingly more vulnerable portion, because it is subjected to the highest tem- resilient Merlot. perature variation, whereas roots tend to be protected by the Vitis sp. presents different rates of deacclimation according soil. We suggest that this is the main reason of the under soil to the stage (early, middle, and late DS). In this way, it has re-growth observed in fields in plants affected by freezing been observed that during the late DS, it is faster than during damage. Finally, we demonstrated that it is possible for a cold the previous stages (Sakai and Larcher 1987; Levitt 1980; deacclimation in a short period of time, and that it depends on Steffen et al. 1989). Thus, temperature fluctuations (including the TA. high TA) during the late winter or extended periods of mild temperatures can lead to deacclimation, making plants vulner- Acknowledgements This work was funded by Agencia Nacional de able to injury from rapid temperature drops (Fennell 2004;Gu Promoción Científica y Tecnológica, Argentina (ANPCyT), CONICET, and PRH 2007 (UNCuyo-AGENCIA). We thank the staff of the Vegetal et al. 2008). Physiology Department and Biological Chemistry of Agronomy Faculty According to our results, plants subjected to high TA, such of UNCuyo Mendoza (especially Bruno Cavagnaro and Emiliano as the case of MZA, probably could not reach the maximum Malovini), Viticulture Laboratory of IAREC in Washington State (espe- capability of acclimation status due small daily cycles of ac- cially Lynn Mills, John Ferguson, and Allan Kawakami), Statistical Department of UNC Córdoba (especially Mónica Balzarini and climation-deacclimation. Moreover, Haynes et al. (1992) Mariano Córdoba), INTA EEA Mendoza Ecophysiology and demonstrated that there is more risk of damage if after a cold Viticulture Department (especially Jorge Perez Peña, Eugenia Galat event, the temperature experiment a rapid increase. A situation Giorgi and Dante Gamboa), IANIGLA CCT-Mendoza (especially like that is common in MZA in late DS when Zonda winds to Federico Gonzalez and Diego Araneo) for sharing their knowledge and technical support and to all the students that participated in our occur and where mean values of TA (16 °C) might experiment research. additional thermal jump of nearly 10 °C, and TA could reach up to 26 °C. Finally, nowadays, it is common to study global climate References change from the GS viewpoint. In fact, Friend et al. (2011) described that spring freeze event can result in substantial Ancapichun S, Garcés-Vargas J (2015) Variability of the Southeast grapevine yield losses by damage in inflorescences and Pacific Subtropical Anticyclone and its impact on sea surface tem- shoots. However, according current predictions, it is expected perature off north-central Chile. Ciencias Mar 41:1–20. doi:10. extreme and abrupt events of warming and freezing waves and 7773/cm.v41i1.2338 Andrews PK, Sandidge CR, Toyama TK (1984) Deep supercooling of Föhn winds during DS (Bernstein et al. 2008) and it could dormant and deacclimating vitis buds. Am J Enol Vitic 35:175–177 induce to a loss of CH (Farrell 2010; Easterling et al. 2000). Battany M, Smith R, Tindula G, Snyder R (2012) Measuring springtime This situation could leave the plants in a more vulnerable temperature inversion conditions for estimating the potential of wind status. In addition, the process of deacclimation occurs at a machines for frost protection. Amer Soc Enology Viticulture 63(3): 456A–456A faster pace than cold acclimation (Kalberera et al. 2006)and Bernstein L, Bosch P, Canziani O, Chen Z, Christ R, Riahi K (2008) plants are in general more rapidly deacclimated during late IPCC, 2007: climate change 2007: synthesis report. Geneva: winter. Thus, the determination of the rate of deacclimation IPCC. ISBN 2-9169-122-4 of vine cultivars is necessary to identify more apt areas for Bianchi AR, Yáñez CE (1992) Las Precipitaciones en el Noroeste Argentino, 2nd edn. INTA EEA, Salta crops. Blennow K (1998) Modelling minimum air temperature in partially and clear felled forests. Agric For Meteorol 91(3):223–235 Brusky Odneal M (1983) Winter bud injury of grapevines 1981–1982. Conclusions Fruit Varieties J 37(2):45–51 Burgos J (2010) Las heladas en la Argentina. Minagri, 2nd edn. Ed. Orientación Gráfica, Buenos Aires The comparison of climatic conditions between MZA and WA Burrows A (1940) The Chinook wind. Or Hist Q 41(1):103–106 revealed contrasting thermal conditions, yet the conditions in Caretta A, Ortega A, Ortíz Maldonado A (2004) Probability of damage to both regions are able to induce freezing damage in vineyards. vines, fruit and olive trees blossom by Zonda wind Mendoza, – One of the most relevant climatic differences between MZA Argentina. Rev FCA UNCuyo 36:49 58 Charrier G, Ameglio T (2011) The timing of leaf fall affects cold accli- and WA is the much higher daily TA in MZA. In addition, the mation by interactions with air temperature through water and car- incidence of Zonda winds during the DS in MZA produces bohydrate contents. Environ Exp Bot 72:351–357. doi:10.1016/j. unseasonal warm spells followed by a freezing event that po- envexpbot.2010.12.019 tentially increases the incidence of cold injury. The absence of Charrier G, Ngao J, Saudreau M, Améglio T (2015) Effects of environmental factors and management practices on microclimate, winter temperature data recorded in the field represents a problem, physiology, and frost resistance in trees. Front Plant Sci 6:1–18. doi: since this study demonstrated that field TA tends to be higher 10.3389/fpls.2015.00259 Int J Biometeorol

Cragin J, Serpe M, Keller M, Shellie K (2017) Dormancy and cold har- Gusta L, Wisniewski M (2013) Understanding plant cold hardiness: an diness transitions in wine grape cultivars chardonnay and cabernet opinion. Physiol Plant 147:4–14. doi:10.1111/j.1399-3054.2012. sauvignon. Am J Enol Vitic. doi:10.5344/ajev.2016.16078 01611.x Dambrorská M (1978) The effect of higher winter temperatures on chang- Hamman RA, Dami IE, Walsh TM, Stushnoff C (1996) Seasonal carbo- es of the frost resistance of grapevine buds. Vitis 17:341–349 hydrate changes and cold hardiness of Chardonnay and Riesling Di Rienzo JA, Guzman AW, Casanoves F (2002) A multiple- grapevines. Am J Enol Vitic 47:31–36 comparisons method based on the distribution of the root node dis- Haynes CL, Lindstrom OM, Dirr MA (1992) Cooling and warming ef- tance of a binary tree. J Agric Biol Environ Stat 7:129–142. doi:10. fects on cold hardiness estimations of three woody ornamental taxa. 1198/10857110260141193 Hortscience 27(12):1308–1309 Di Rienzo, JA, Casanoves, F, Balzarini, MG, Gonzalez L, Tablada M, Howell GS, Shaulis N (1980) Factors influencing within-vine variation in Robledo, CW (2016). InfoStat versión 2016. Grupo InfoStat, FCA, the cold resistance of cane and primary bud tissues. Am J Enol Vitic Universidad Nacional de Córdoba, Argentina. http://www.infostat. 31(2):158–161 com.ar Hubackova M (1996) Dependence of grapevine bud cold hardiness on Eagles CF (1989) Temperature-induced changes in cold tolerance of fluctuations in winter temperatures. Am J Enol Vitic 47:100–102 Lolium perenne. J Agric Sci 113:339–347. doi:10.1017/ Jiang H, Howell GS (2002) Correlation and regression analyses of cold S0021859600070027 hardiness, air temperatures, and water content of Concord grape- Easterling D, Meehl G, Parmesan C, Changnon S, Karl T, Mearns L vines. Am J Enol Vitic 53:227–230 (2000) Climate extremes: observations, modeling and impacts. Jordan D, Smith W (1995) Microclimate factors influencing the frequen- Science 289(5487):2068–2074. doi:10.1126/science.289.5487. cy and duration of growth season frost for subalpine plants. Agric 2068 For Meteorol 77(1):17–30 Echarte M, Angeloni P, Jaimes F, Tognetti J, Izquierdo N, Valentinuz O, Kalberera S, Wisniewski M, Arora R (2006) Deacclimation and Aguirrezába L (2010) Night temperature and intercepted solar radi- reacclimation of cold-hardy plants: current understanding and ation additively contribute to oleic acid percentage in sunflower oil. emerging concepts. Plant Sci 171(1):3–16. doi:10.1016/j.plantsci. Field Crops Res 119(1):27–35. doi:10.1016/j.fcr.2010.06.011 2006.02.013 Fall S, Watts A, Nielsen-Gammon J, Jones E, Niyogi D, Christy J, Pielke Keller M, Mills LJ (2007) Effect of pruning on recovery and productivity R (2011) Analysis of the impacts of station exposure on the U.S. of cold-injured merlot grapevines. Am J Enol Vitic 58(3):351–357 Historical Climatology Network temperatures and temperature Leuning R, Cremer KW (1988) Leaf temperatures during radiation frost trends. J Geophys Res 116:D14120. doi:10.1029/2010JD015146 part I. Observations. Agric For Meteorol 42(2–3):121–133 Farrell J (2010) Deacclimation response to winter temperature fluctua- Levitt J (1980) Responses of plants to environmental stresses, Ed 2 edn. tions in Lolium perenne. Graduate Theses and Dissertations. Paper Academic Press, New York 11831. Iowa State University Mass C, Dotson B (2010) Major extratropical cyclones of the Northwest Fennell A (2004) Freezing tolerance and injury in grapevines. J Crop United States: historical review, climatology, and synoptic environ- Improv 10(1–2):201–235. doi:10.1300/J411v10n01 ment. Mon Weather Rev 138(7):2499–2527. doi:10.1175/ Ferguson J, Moyer M, Mills J, Hoogenboom G, Keller M (2014) 2010MWR3213.1 Modeling dormant bud cold hardiness and budbreak in twenty- Paroschy JH, Meiering AG, Peterson RL, Hostetter G, Neff A (1980) three Vitis genotypes reveals variation by region of origin. Am J Mechanical winter injury in grapevine trunks. Am J Enol Vitic Enol Vitic 65(1):59–71. doi:10.5344/ajev.2013.13098 31(3):227–232 Ferguson J, Tarara J, Mills L, Grove G, Keller M (2011) Dynamic thermal Menne M, Williams C, Palecki M (2010) On the reliability of the U.S. time model of cold hardiness for dormant grapevine buds. Ann Bot surface temperature record. J Geophys Res 115:D11108. doi:10. Lond 107(3):389–396. doi:10.1093/aob/mcq263 1029/2009JD013094 Friend AP, Trought MCT, Stushnoff C, Wells GH (2011) Effect of Mills LJ, Ferguson JC, Keller M (2006) Cold-hardiness evaluation of delaying budburst on shoot development and yield of Vitis vinifera grapevine buds and cane tissues. Am J Enol Vitic 57:194–200 L. Chardonnay ‘Mendoza’ after a spring freeze event. Aust J Grape Mullins MG, Bouquet A, Williams LE (1992) Biology of the grapevine. Wine Res 17:378–382. doi:10.1111/j.1755-0238.2011.00162.x Cambridge University Press, Cambridge Fuchigami LH, Weiser CJ, Kobayashi KD, Timmis R, Gusta V (1982) A Norte F, Simonelli S (2016) Validation of a statistical forecast model for degree growth stage (GS) model and cold acclimation in temperate Zonda wind in West Argentina based on the vertical atmospheric woody plants. In: Li PH, Sakai A (eds) Plant cold hardiness and structure. Atmos Clim Sci 6(1):35–50 freezing stress. II Plant acclimation. Academic Press, New York, Okamoto G, Shippin W, Ken H (2000) Cold resistance in root and cane of pp 93–116 own-root ‘Kyoho’ grapevines. Sci Rep Fac Agric Okayama Univ Gedalof ZE, Peterson DL, Mantua NJ (2005) Atmospheric, climatic, and 89:23–29 ecological controls on extreme wildfire years in the northwestern Pagter M, Arora R (2013) Winter survival and deacclimation of peren- United States. Ecol Appl 15(1):154–174 nials under warming climate: physiological perspectives. Physiol Giantomasi M, Roig Junent F, Villagra P, Srur A (2009) Annual variation Plant 147:75–87. doi:10.1111/j.1399-3054.2012.01650.x and influence of climate on the ring width and wood hydrosystem of Pagter M, Williams M (2011) Frost dehardening and rehardening of Prosopis flexuosa DC trees using image analysis. Trees 23(1):117– Hydrangea macrophylla stems and buds. Hortscience 46(8):1121– 126. doi:10.1007/s00468-008-0260-5 1126 Goffinet M (2004) Anatomy of grapevine winter injury and recover. Proebsting EL, Ahmedullah M, Brummund V (1980) Seasonal changes Cornell University Department of Horticultural Sciences, NY State in low temperature resistance of grape buds. Am J Enol Vitic 31: Agricultural Experiment Station, Geneva 329–336 Gu L, Hanson P, Mac Post W, Kaiser D, Yang B, Nemani R, Meyers T Core Team R (2013) R: a language and environment for statistical com- (2008) The 2007 eastern US spring freeze: increased cold damage in puting. R Foundation for Statistical Computing, Vienna http://www. a warming world. Bioscience 58(3):2532–26 R-project.org/ Gu S, Ding P, Howard S (2002) Effect of temperature and exposure time Renaud V, Rebetez M (2009) Comparison between open-site and below- on cold hardiness of primary buds during the dormant season in canopy climatic conditions in Switzerland during the exceptionally ‘Concord’, ‘Norton’, ‘Vignoles’ and ‘St. Vincent’ grapevines. J hot summer of 2003. Agric For Meteorol 149(5):873–880. doi:10. Hortic Sci Biotechnol 77:635–639 1016/j.agrformet.2008.11.006 Int J Biometeorol

Robinson J (2006) The Oxford companion to wine, 3rd edn. Press, 4th Intl. Symp.Cool Climate Viticulture and Enology. New York Oxford University State Agricultural Experiment Station, Geneva, pp 23–28 Rubio S, Dantas D, Bressan-Smith R, Pérez F (2016) Relationship Wisniewski M, Wolf T, Fuchigami L (1996) Biochemical and biophysical between endodormancy and cold hardiness in grapevine buds. mechanisms of cold hardiness in woody plants. In: Henick-Kling T, J Plant Growth Regul 35(1):266–275. doi:10.1007/s00344- Wolf TE, Harkness EM (eds) Proc. 4th Intl. Symp.Cool Climate 015-9531-8 Viticulture and Enology. New York State Agricultural Experiment Sage R, Kubien D (2007) The temperature response of C3 and C4 pho- Station, Geneva, pp 14–22 tosynthesis. Plant Cell Environ 30:1086–1106. doi:10.1111/j.1365- Wolf TK, Pool RM (1987) Factors affecting exotherm detection in the 3040.2007.01682.x differential thermal analysis of grapevine dormant buds. J Amer Soc Sakai A, Larcher W (1987) Frost survival of plants. Ecological Studies Hort Sci 112:520–525 62. Springer-Verlag, Berlin Wolf TK, Cook MK (1992) Seasonal deacclimation patterns of three Steffen K, Arora R, Palta J (1989) Relative sensitivity of photosynthesis grape cultivars at constant, warm temperature. Am J Enol Vitic 43: and respiration to freeze-thaw stress in herbaceous species impor- 171–179 – tance of realistic freeze-thaw protocols. Plant Physiol 89(4):1372 WRCC (2013) Western Regional Climate Center, Cooperative 1379 Climatological Data Summaries http://www.wrcc.dri.edu/ Tarara JM, Hoheisel GA (2007) Low-cost shielding to minimize radiation climatedata/climsum/. Accessed 4 January 2017 errors of temperature sensors in the field. Hortscience 42(6):1372– Xin Z, Browse J (2000) Cold comfort farm: the acclimation of plants to 1379 freezing temperatures. Plant Cell Environ 23:893–902 Thompson K, Grime J, Mason G (1977) Seed germination in response to diurnal fluctuations of temperature. Nature 267(5607):147–149 Zabadal TJ, Dami IE, Goffinet MC, Martinson T, Chien M (2007) Winter injury to grapevines and methods of protection. Mich State Univ Ext Wample R, Wolf T (1996) Practical considerations that impact vine cold – hardiness. In: Henick-Kling T, Wolf TE, Harkness EM (eds) Proc. E2930:1 44. doi:10.1016/j.cropro.2005.11.004